Difference test of paper ashes by XRF combined with multivariate statistical analysis

JIANG Hong; WANG Xin; XU Lele; ZHU Xiaohan

doi:10.7510/jgjs.issn.1001-3806.2021.03.009

In order to quickly and accurately test and identify the residual paper ashes on the scene of the case, 32 paper ashes from different sources and uses were determined by X-ray fluorescence spectrometry combined with system clustering and principal component analysis of multivariate statistics. Theoretical analysis and experimental verification were then carried out. The results show that the samples can be distinguished accurately according to the types and contents of the elements in the samples. At the same time, 32 samples are divided into five categories, and the classification results of the two methods are basically the same. The cumulative variance contribution rate of PC₂ is as high as 99.86%. This research method does not destroy the sample, has good reproducibility, and the results are scientific and ideal, which provides a theoretical basis for the examination of trace evidence.

HTML

引言

对犯罪现场遗留的纸张及其灰烬的分析检验是法庭科学研究的重要课题之一。纸张灰烬物证常见于各类案件现场，犯罪嫌疑人常用此手段隐藏犯罪信息^[1]。在公安实践应用中，系统科学地分析鉴别案件现场残留的各种纸张灰烬，对于火灾原因的认定、提供案件侦破线索等有重要意义^[2]。因此，对纸张灰烬的鉴别检验，有益于侦查人员确定下一步侦查方向和范围，为案件分析提供思路。目前，对纸张灰烬的检验方法主要有扫描电镜/能谱法^[2-4]等。X射线荧光光谱法(X-ray fluorescence spectrometer，XRF)因其灵敏度高、分析试样不受破坏、能实现多元素同时测定等优点广泛应用于物证鉴定^[5-7]。多元统计学是化学计量常用方法，包括回归分析、聚类分析、主成分分析等，借助多元统计分析，可实现数据降维、分组以及分类，挖掘多变量的数据之间的关系，实现对样本的有效归类^[8-10]。

本实验中应用X射线荧光光谱，对32个不同来源和种类的纸张灰烬样品中的无机元素进行定性和半定量分析，以获取原纸张所含无机元素的种类和含量，并借助统计产品与服务解决方案(statistical product and service solutions, SPSS)软件，采用系统聚类分析和主成分分析对实验数据进行处理和验证，从多变量的数据中提取信息和规律，得到了较为满意的结果。

1. 实验

1.1. 实验仪器及条件

X-MET8000X射线荧光光谱仪(牛津仪器)，Rh靶，电压为45kV，电流为40μA，功率为1.8kW，测试时间为80s。

1.2. 实验样品

不同来源和种类的纸张灰烬样品共32个(部分样品见表 1)。

sample number	type	source	specification (cm×cm)
1	napkin	Vinda	13.5×19.5
2	worksheet	Yurongfeng notebook	20×14
4	worksheet	PPSUC scratch paper	26×18
6	shoebox paper	BASTO
8	xuan paper	copybook
9	copy paper	copybook
10	postcard	PPSUC postcard	15×10
12	cardboard	Lianhua Qingwen capsule box	20×14
13	cardboard	SITILON facial mask box	32×16
14	cardboard	Franzzi cookie box	25×19
17	plastic tissue	Via lemon tea	20×12
18	cardboard	DHC lipstick packaging box	18×6
19	cardboard	Mac lipstick packaging box	8×6
20	cardboard	Zhizhu sock box	16×6
21	cardboard	auberge bracelet box	22×8
22	newspaper	PPSUC souvenir	20×14
25	description paper	auberge description	25×8
26	thermal paper	PPSUC supermarket invoice	22×5
27	cardboard	fruit knife packing box	18×5
28	business card	Doraemon’s pocket	9×5
29	express details	ZTO express	10×10
30	oil surface sticker	stickers	10×8
31	oil paper	pin-up picture	22×18
32	corrugated paper	express box

Table 1. Partial sample table of paper ash samples

1.3. 实验方法

取一定量样品置于干锅中点燃，完全燃烧后，待自然冷却后，将灰烬样品混合均匀，移至洁净样品塑料包装袋上，待测。在上述实验条件下，分别对32个样品依次进行测定，每个样品测3次取平均值。

3. 结论

实验表明，利用X射线荧光光谱法对纸张灰烬样品进行定量与定性分析，根据特征元素的种类和含量可将32个不同来源和用途的纸张灰烬样品进行准确区分；并从多角度对数据进行了多元统计分析，充分挖掘了变量之间的关系，提高了对纸张灰烬样品分析分类的科学性和准确性，可以获得理想的聚类效果。本实验中的方法操作方便，实验结果可靠，且对样品无损，可以准确地通过纸张灰烬样品区分原纸张，可应用于公安实际办案。

Reference (20)

[1]	CHEN J. Research on the characteristics of corrugated paper ignition experiment and residue trace[C]//2015 Fire Technology and Engineering Academic Conference Proceedings. Beijing: China Fire Protection Association, 2015: 352-354(in Chinese).
[2]	HUANG Ch, SHI W, LI J H. Study on the identification of the burning ash of paper materials by SEM energy spectrum[J]. China Papermaking, 2016, 35(9): 31-37.
[3]	HUANG Ch. Energy spectrum identification of common paper material ashes by electron microscopy[C]//Proceedings of 2015 National Conference on Electron Microscopy. Chengdu: Chinese Physical Society, 2015: 108-109(in Chinese).
[4]	HUANG Ch, LI J H. Study on the characteristics of combustion residues by SEM energy dispersive spectrometer[J]. Fire Science and Technology, 2014, 33(8): 965-968.
[5]	CHEN Zh, JIANG H, LI Ch Y. A study on the inspection of disposable paper cups by X-ray fluorescence spectrometry[J]. ZhongHua Paper, 2018, 39(22): 32-36.
[6]	SHI X X, ZHENG Ch Y, FANG Q Y. Progress in the application of XRF in food, biomedicine and cosmetics[J]. Analytical Instruments, 2019, 222(1): 12-17.
[7]	SU H F, LIU Ch, HU S L. Progress in the application of microbeam X-ray fluorescence spectrometry in forensic identification[J]. Journal of Forensic Medicine, 2013, 29(1): 43-48.
[8]	FILZMOSER P. Introduction to multivariate statistical analysis in chemometrics[J]. Applied Spectroscopy, 2010, 64(4): 112-112.
[9]	SENA M M, ALMEIDA M R, BRAGA J W B, et al. Multivariate statistical analysis and chemometrics[M]. Boca Raton, USA: CRC Press, 2017: 273-314.
[10]	MENG Q L, ZHANG Y, SHANG J. Nondestructive detection of apple surface scar by optical fiber spectroscopy combined with pattern re-cognition[J]. Laser Technology, 2019, 43(5): 676-680.
[11]	LIU Zh H. Committed to resource-based utilization and leading the production of light calcium carbonate for paper-making[J]. Papermaking Information, 2017, 32(11): 59-62.
[12]	ZHAI R, ZHAO H F, ZHOU X F. Study on the preparation of wet strength paper by sodium hydroxide urea zinc oxide water system[J]. Cellulose Science and Technology, 2019, 27(3): 30-35.
[13]	YANG W H, HUANG Q R, ZHAO T Y. Research and application of a composite paper fungicide[J]. China Paper, 2008, 27(12): 36-38.
[14]	CHEN X F, XU Y, LIU W. Study on the properties of composite titanium dioxide and its application in decorative base paper[J]. China Paper, 2015, 34(12): 1-6.
[15]	XUE M G, WANG Sh F, HUANG Ch X. Determination and source analysis of Pb, Cd, Cr and Hg in printed paper food packaging materials[J]. Journal of Chemical Industry, 2010, 61(12): 3258-3265.
[16]	SINGH P, SRIVASTAVA N, SINGH P. Analysis of various iron nanoparticles and compounds in pulp & paper mill waste water treatment[J]. Integrated Research Advances, 2017, 4(2): 24-28.
[17]	ZHANG W T. SPSS advanced course of statistical analysis[M]. Beijing: Higher Education Press, 2012: 260-263(in Chinese).
[18]	CHU X L. Chemometrics and molecular light spectroscopy[M]. Beijing: Chemical Industry Press, 2011: 55-57(in Chinese).
[19]	HU L F. Five commonly used cluster analysis methods and their comparison[J]. Zhejiang Statistics, 2007, 14(4): 11-13.
[20]	FU D Y. Statistical test in principal component analysis[J]. Statistical Education, 2007, 14(9): 4-7.

sample number	Ca	Cl	Fe	Zn	Ti	Pb	Br
1	305	457	126	23469	0	71	1191
2	260	417	224	15184	0	42	0
3	707	0	124	13754	0	51	3572
4	19374	1469	244	16342	23	31	0
5	8618	0	189	13755	399	33	21
6	435	119	175	21978	27	72	972
7	102	0	153	18539	0	48	24
8	198	357	167	18755	30	63	1799
9	338	688	171	37209	20	0	0
10	782	104	155	11007	0	49	338
11	340	0	160	20904	0	87	812
12	205	386	126	11108	0	46	0
13	460	476	183	11533	16	61	31
14	707	449	154	13097	0	47	170
15	270	0	146	11579	0	69	0
16	283	0	133	21284	13	59	1493
17	282	1108	1151	23250	23	0	0
18	1587	190	152	10812	44	27	0
19	92	206	141	22345	11	48	941
20	1020	2469	194	22961	15	46	0
21	256	354	190	13647	0	40	0
22	295	123	135	22727	0	61	1014
25	360	414	164	18949	13	0	0
26	404	469	129	20601	0	67	469
27	420	354	198	21321	15	79	1023
28	319	534	219	24399	0	58	401
29	615	123	172	17307	14	45	0
30	316	440	142	29952	0	0	0
31	688	95	201	13237	0	33	0
32	8340	1427	252	59	0	0	0

sample number	Zn/Ca	Br/Cl
6	50.52	5.55
8	94.72	5.03
13	25.07	0.07
19	242.88	4.57
27	50.76	2.89

sample number	PC₁	PC₂	sample number	PC₁	PC₂	sample number	PC₁	PC₂
1	0.045	-0.025	13	0.045	-0.005	25	0.045	-0.017
2	0.045	-0.019	14	0.045	0.003	26	0.045	-0.018
4	0.028	0.476	17	0.045	-0.023	27	0.045	-0.021
6	0.045	-0.020	18	0.045	0.061	28	0.045	-0.022
8	0.045	-0.030	19	0.045	-0.029	29	0.045	-0.007
9	0.045	-0.025	20	0.045	0.000	30	0.045	-0.024
10	0.045	0.012	21	0.045	-0.018	31	0.045	0.003
12	0.045	-0.017	22	0.045	-0.024	32	-0.003	0.616

Difference test of paper ashes by XRF combined with multivariate statistical analysis

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Proportional views

Difference test of paper ashes by XRF combined with multivariate statistical analysis